Iaroslav Bespalov

Ivan Zakazov, Vladimir Shaposhnikov, Iaroslav Bespalov et al.

Generalizability of deep learning models may be severely affected by the difference in the distributions of the train (source domain) and the test (target domain) sets, e.g., when the sets are produced by different hardware. As a consequence of this domain shift, a certain model might perform well on data from one clinic, and then fail when deployed in another. We propose a very light and transparent approach to perform test-time domain adaptation. The idea is to substitute the target low-frequency Fourier space components that are deemed to reflect the style of an image. To maximize the performance, we implement the "optimal style donor" selection technique, and use a number of source data points for altering a single target scan appearance (Multi-Source Transferring). We study the effect of severity of domain shift on the performance of the method, and show that our training-free approach reaches the state-of-the-art level of complicated deep domain adaptation models. The code for our experiments is released.

Vladimir Shaposhnikov, Aleksandr Nesterov, Ilia Kopanichuk et al.

We present CLARITY (Clinical Assistant for Routing, Inference and Triage), an AI-driven platform designed to facilitate patient-to-specialist routing, clinical consultations, and severity assessment of patient conditions. Its hybrid architecture combines a Finite State Machine (FSM) for structured dialogue flows with collaborative agents that employ Large Language Model (LLM) to analyze symptoms and prioritize referrals to appropriate specialists. Built on a modular microservices framework, CLARITY ensures safe, efficient, and robust performance, flexible and readily scalable to meet the demands of existing workflows and IT solutions in healthcare. We report integration of our clinical assistant into a large-scale national interhospital platform, with more than 55,000 content-rich user dialogues completed within the two months of deployment, 2,500 of which were expert-annotated for subsequent validation. The validation results show that CLARITY surpasses human-level performance in terms of the first-attempt routing precision, naturally requiring up to 3 times shorter duration of the consultation than with a human.

Ilia Kopanichuk, Petr Anokhin, Vladimir Shaposhnikov et al.

The integration of artificial intelligence (AI) into medical diagnostic workflows requires robust and consistent evaluation methods to ensure reliability, clinical relevance, and the inherent variability in expert judgments. Traditional metrics like precision and recall often fail to account for the inherent variability in expert judgments, leading to inconsistent assessments of AI performance. Inter-rater agreement statistics like Cohen's Kappa are more reliable but they lack interpretability. We introduce Relative Precision and Recall of Algorithmic Diagnostics (RPAD and RRAD) - a new evaluation metrics that compare AI outputs against multiple expert opinions rather than a single reference. By normalizing performance against inter-expert disagreement, these metrics provide a more stable and realistic measure of the quality of predicted diagnosis. In addition to the comprehensive analysis of diagnostic quality measures, our study contains a very important side result. Our evaluation methodology allows us to avoid selecting diagnoses from a limited list when evaluating a given case. Instead, both the models being tested and the examiners verifying them arrive at a free-form diagnosis. In this automated methodology for establishing the identity of free-form clinical diagnoses, a remarkable 98% accuracy becomes attainable. We evaluate our approach using 360 medical dialogues, comparing multiple large language models (LLMs) against a panel of physicians. Large-scale study shows that top-performing models, such as DeepSeek-V3, achieve consistency on par with or exceeding expert consensus. Moreover, we demonstrate that expert judgments exhibit significant variability - often greater than that between AI and humans. This finding underscores the limitations of any absolute metrics and supports the need to adopt relative metrics in medical AI.

Ivan Bakulin, Ilia Kopanichuk, Iaroslav Bespalov et al.

The rapid advancement of Large Language Models (LLMs) has introduced significant challenges in moderating user-model interactions. While LLMs demonstrate remarkable capabilities, they remain vulnerable to adversarial attacks, particularly ``jailbreaking'' techniques that bypass content safety measures. Current content moderation systems, which primarily rely on input prompt filtering, have proven insufficient, with techniques like Best-of-N (BoN) jailbreaking achieving success rates of 80% or more against popular LLMs. In this paper, we introduce Flexible LLM-Assisted Moderation Engine (FLAME): a new approach that shifts the focus from input filtering to output moderation. Unlike traditional circuit-breaking methods that analyze user queries, FLAME evaluates model responses, offering several key advantages: (1) computational efficiency in both training and inference, (2) enhanced resistance to BoN jailbreaking attacks, and (3) flexibility in defining and updating safety criteria through customizable topic filtering. Our experiments demonstrate that FLAME significantly outperforms current moderation systems. For example, FLAME reduces attack success rate in GPT-4o-mini and DeepSeek-v3 by a factor of ~9, while maintaining low computational overhead. We provide comprehensive evaluation on various LLMs and analyze the engine's efficiency against the state-of-the-art jailbreaking. This work contributes to the development of more robust and adaptable content moderation systems for LLMs.

Iaroslav Bespalov, Nazar Buzun, Oleg Kachan et al.

The training of Generative Adversarial Networks (GANs) requires a large amount of data, stimulating the development of new augmentation methods to alleviate the challenge. Oftentimes, these methods either fail to produce enough new data or expand the dataset beyond the original manifold. In this paper, we propose a new augmentation method that guarantees to keep the new data within the original data manifold thanks to the optimal transport theory. The proposed algorithm finds cliques in the nearest-neighbors graph and, at each sampling iteration, randomly draws one clique to compute the Wasserstein barycenter with random uniform weights. These barycenters then become the new natural-looking elements that one could add to the dataset. We apply this approach to the problem of landmarks detection and augment the available annotation in both unpaired and in semi-supervised scenarios. Additionally, the idea is validated on cardiac data for the task of medical segmentation. Our approach reduces the overfitting and improves the quality metrics beyond the original data outcome and beyond the result obtained with popular modern augmentation methods.

Viktor Shipitsin, Iaroslav Bespalov, Dmitry V. Dylov

We devise a universal adaptive neural layer to "learn" optimal frequency filter for each image together with the weights of the base neural network that performs some computer vision task. The proposed approach takes the source image in the spatial domain, automatically selects the best frequencies from the frequency domain, and transmits the inverse-transform image to the main neural network. Remarkably, such a simple add-on layer dramatically improves the performance of the main network regardless of its design. We observe that the light networks gain a noticeable boost in the performance metrics; whereas, the training of the heavy ones converges faster when our adaptive layer is allowed to "learn" alongside the main architecture. We validate the idea in four classical computer vision tasks: classification, segmentation, denoising, and erasing, considering popular natural and medical data benchmarks.

Iaroslav Bespalov, Nazar Buzun, Dmitry V. Dylov

Unsupervised retrieval of image features is vital for many computer vision tasks where the annotation is missing or scarce. In this work, we propose a new unsupervised approach to detect the landmarks in images, validating it on the popular task of human face key-points extraction. The method is based on the idea of auto-encoding the wanted landmarks in the latent space while discarding the non-essential information (and effectively preserving the interpretability). The interpretable latent space representation (the bottleneck containing nothing but the wanted key-points) is achieved by a new two-step regularization approach. The first regularization step evaluates transport distance from a given set of landmarks to some average value (the barycenter by Wasserstein distance). The second regularization step controls deviations from the barycenter by applying random geometric deformations synchronously to the initial image and to the encoded landmarks. We demonstrate the effectiveness of the approach both in unsupervised and semi-supervised training scenarios using 300-W, CelebA, and MAFL datasets. The proposed regularization paradigm is shown to prevent overfitting, and the detection quality is shown to improve beyond the state-of-the-art face models.